Google launched a new conversational agent (chatbot), Bard, and gave access to this experimental technology to a limited number of people in the United States and Great Britain.
Bard competes with similar technologies from Microsoft and its partner, San Francisco-based startup OpenAI. Google was cautious in releasing it, however, as it attempts to control unexpected behavior from this type of technology. It deploys this tool as a service operating separately from its internet search engine and other products.
Here is a preview of the new chatbot.
Bard tells you right away that he makes mistakes. “I am Bard, your creative and helpful collaborator. I have limitations and won’t always be able to achieve them, but your feedback will help me improve,” reads a message at the top of the page.
Like OpenAI’s ChatGPT and Microsoft’s version adapted for Bing, Bard is based on a large language model, or LLM, a type of artificial intelligence technology that learns by analyzing large amounts of data from the internet. This means that he sometimes makes mistakes and can make things up.
To get started, Bard offers a few questions, including “Explain why large language models sometimes make mistakes.”
Bard is not a search engine. This is an experimental system intended to show people how they can use this type of chatbot.
It can generate ideas. He can write blog posts. And he can answer questions with facts or opinions.
To the question “What is the most important moment in the history of the United States?” “, he answers :
The most important moment in American history is a matter of opinion, but some of the most frequently cited candidates include:
– The American Revolution (1775-1783): The founding of the United States of America and the establishment of a new nation based on the principles of freedom and democracy.
– The Civil War (1861-1865): The war that ended slavery in the United States and preserved the Union.
– The Great Depression (1929-1939): The economic crisis that plunged the United States into poverty and unemployment…
The list continues with other events.
Like other similar technologies, Bard generates new text each time you enter a query. He can answer the same question in different ways.
Asked again about the most important moment in American history, Bard replied that it was undoubtedly the American Revolution.
Like Microsoft’s Bing chatbot and other similar technologies developed by startups like You.com and Proximity, Bard sometimes annotates his answers so you can see his sources. It also connects to Google’s extensive website index, so its responses can include the most up-to-date information published on the internet.
But that doesn’t always lead to better sources. When the chatbot wrote that the most important moment in American history was the American Revolution, he cited a blog, Pix Style Me, written in a mix of English and Chinese and adorned with cartoon cats .
When asked why he cited that particular source, the robot insisted that he cited Wikipedia.
While using OpenAI’s latest version of ChatGPT this month, AI researcher and professor Oren Etzioni asked the chatbot, “What is the relationship between Oren Etzioni and Eli Etzioni?” the robot answered correctly that Oren and Eli are father and son.
When he asked Bard the same question, he refused to answer. “My knowledge of this person is limited. Is there anything else I can do to help you meet this request? »
Eli Collins, vice president of research at Google, said the bot often refuses to answer about specific people because it might generate misinformation about them – a phenomenon AI researchers call “hallucination. “.
Chatbots often hallucinate internet addresses. When Bard was asked to provide several websites discussing the latest cancer research, he declined to do so.
ChatGPT will respond to similar requests (and, yes, it will invent websites). Mr. Collins said Google Bard tends to avoid providing medical, legal or financial advice because it could lead to misinformation.
